Patterns of Physical Activity and Self-rated Health Among Adult Populations in South Asia

Introduction: Although South Asians are considered to be at high risk for cardiovascular diseases, research evidence on the health impacts of physical activity (PA) remains very limited. In this study we aimed to explore the patterns of PA and to investigate whether engaging in regular PA is associated with better Self-Rated Health (SRH) among South Asians. Methods: Cross-sectional data on population health were drawn from the World Health Survey of WHO. Subjects were 28,020 male and female South Asians (from Bangladesh, India, Nepal, and Sri Lanka) aged 18 years and above. Data were analysed using descriptive and multivariable logistic regression analyses. Results: The proportion of the sample population reported good SRH was 44.3%, 58.7%, 37.7%, and 73.7% in Bangladeshis, Indians, Nepalese, and Sri Lankans, respectively. Regular engagement in moderate PA was highest in Nepal (69.7%) and lowest in Bangladesh (37.4%). Vigorous PA was highest in India (29.9%) and lowest in Bangladesh (17.9%). In Bangladesh, compared to those never engaged in MPA, those who engaged for 1-2, 3-4, 5-6, or 7 days a week were 30% [AOR=1.306; 95%CI 1.085-1.572], 33% [AOR=1.326; 95%CI 1.093-1.609], 39% [AOR=1.389; 95%CI 1.125-1.716], and 46% [AOR=1.459; 95%CI 1.249-1.705] more likely to report being in good health, respectively. Conclusions: We found that self-reported engagement in physical activities varies in South Asian countries. Since engaging in PA may help improve subjective and objective health status, health policy makers need to focus on designing exercise-friendly neighbourhoods in an attempt to promote population health.


Introduction
The construct of self-rated health (SRH) is an inclusive measure of public health, and it is used as a reliable predictor of quality of life, subjective well-being, disability, morbidity, and mortality 1,2 . SRH is one of the most widely used predictors of health risk and prognosis compared with/to other objective measures 3 . Possible explanations for the efficacy of self-assessments of health include its multifaceted representation of an individual's general perception of health, including biological, psychosocial, and cultural dimensions of health and expressiveness 4 . SRH has also been found to be related to clinical measures of health 3 , and it was proposed that general practitioners can utilize SRH measures in clinical encounters 5 . Moreover, SRH can be regarded as a more inclusive measure of health status than clinical diagnosis, as it tends to be sensitive to social determinants of health such as education, socioeconomic status, and living conditions, which have direct influences on health and well-being and on shaping individuals perception of health and illness 4,5 . SRH is therefore able to provide information above and beyond typical clinical evaluation and thus offers a comprehensive way of assessing a patients' overall health status 4,6 .
In recent years, there has been an increasing research interest on the impact of various lifestyle factors and health related behaviours on SRH 7,8 . Behavioural aspects, such as tobacco smoking, alcohol drinking, dietary habits, and engaging in physical activity (PA) are explored in relation to how they correlate with SRH among people of different age groups and socioeconomic backgrounds. Physical inactivity is regarded as a growing public health issue both in developed and developing countries. In Europe and other industrialised societies, increasing sedentary lifestyle has been shown to be associated with worse health and all-cause mortality, independent of level of PA [9][10][11] . According to WHO, public health burden of physical inactivity is high and causes an estimated 600,000 deaths per year in Europe alone 11 . Worldwide, physical inactivity is responsible for 6% of the burden of coronary heart disease, 7% of type 2 diabetes, and 10% of breast and colon cancer 12 . It has also been identified as the fourth leading risk factor for global mortality in 2010, accounting for roughly 13.4 million disability adjusted life years (DALYs) 15 and 6% of all deaths 13 .
There is a growing consensus that moderate-to vigorous-intensity PA has a key preventive role in noncommunicable diseases (NCDs), including obesity, cardiovascular disease, type-2 diabetes, and some cancers 9,10 . Lack of PA during adolescence was reported to be a significant predictor of abdominal obesity in young adulthood leading to a self-perpetuating vicious circle of obesity and physical inactivity 14 . In addition to its contribution to increased morbidity and mortality, physical inactivity is also responsible for a substantial economic burden. Epidemiological evidence on the role of PA on SRH is necessary for making informed health policies that can promote PA in the general population.
South Asians are people who identify with the cultures of Bangladesh, India, Sri Lanka, and Nepal, and account for about a quarter of the global population with a unique epidemiological and sociodemographic profile. Though South Asians are considered at-risk population for cardiovascular diseases, research evidence on the health impacts of PA in this population remains very limited. Therefore, epidemiological evidence from other regions may not be applicable for the population in this region. This study was carried out to provide insights on the pattern of PA in South Asians, and to investigate the association between frequency of PA and SRH among the adult population. Data used in this study were extracted from the World Health Survey program of WHO conducted during 2002-2004 that included four South Asian countries: Bangladesh, India, Nepal, and Sri Lanka.

Data source
This study was based on the data extracted from WHO World Health Survey conducted between 2002 and 2004, available from WHO upon request. Objectives of the WHO-funded survey were to provide reliable and nationally comparable data on a wide range of health and socioeconomic indicators to facilitate evidence-based health policy making. These data are utilized by many researchers due to lack of more recent data on health behaviour and self-rated health in South Asian population. The program is operational in 70 countries including four South Asian countries: Bangladesh, India, Nepal, and Sri Lanka. Further details regarding the original survey study are published elsewhere 16 .

Variables of interest
Self-rated health status was the outcome variable in this study. Emerging evidence suggests predictability of SRH for both non-clinical and clinical outcomes, and it is being proposed to family physicians as an efficient yet simple way for therapeutic decision making 28,29 . Respondents were asked to rate their health on a scale from 1 to 5 with the following response This journal is published by theUniversity Library System of the University of Pittsburgh as part of its D-Scribe Digital Publishing Program and is cosponsored by the University of Pittsburgh Press.
Central Asian Journal of Global Health  30 . The validity of the single-item tool to measure subjective health was mentioned in previous studies 31,32 .
The predictor variable of primary interest was PA. The two types of PA used in this study were moderate PA (MPA) and vigorous PA (VPA). VPA was measured by the following question: "Vigorous activities make you breathe much harder than normal and may include heavy lifting, digging, aerobics, or fast bicycling. Think only about those physical activities that you performed for at least 10 minutes at a time. During the last 7 days, on how many days did you do vigorous physical activities?" MPA was measured by the following question: "Moderate physical activities make you breathe somewhat harder than normal and may include carrying light loads, bicycling at a regular pace, or doubles tennis. Do not include walking. Again, think about only those physical activities that you performed for at least 10 minutes at a time. During the last 7 days, on how many days did you do moderate physical activities?" Answers ranged from 0 to 7 days and were categorised as follows: 0 days (never), 1-2 days, 3-4 days, 5-6 days, and every day.

Ethical considerations
Informed Consent was obtained from the recruited participants before their participation in the survey. Participation was completely voluntary, and the respondent had the choice to refuse to take part in the interview. The data used in this study were secondary,

Statistical analysis
Datasets were checked for missing values and outliers. Data were cleaned to retain the maximum number of observations. Sample characteristics were analysed through univariate analysis. Cross tabulation was used to measure the distribution of the sociodemographic variables across the outcome SRH variable. Chi-square tests were conducted to assess the group differences for Good vs. Poor self-rated health. Variables that had a p-value below 0.25 were entered into the final regression analysis 17 . Four separate regression models were run for each country. The outcomes of the regression (binary logistic) analyses were reported in terms of adjusted odds ratios (AOR) and corresponding 95% confidence intervals. All analyses were performed with SPSS version 22.

Results
Descriptive sample characteristics were provided in Table 1. In short, the mean age was highest in Sri Lankans (40.78, SD 15.22) and lowest in Bangladeshis (38.47, SD 14.81). The majority of the participants were between 18 and 29 years of age, female, and currently married. Rate of literacy was highest for Sri Lanka (94.4%) and lowest in Nepal (40%). However, the rate of pre-university/university level education was highest in India (10.1%), followed by Bangladesh (4.3%) and Sri Lanka (2.4%). Regular engagement in MPA was highest for Nepal (69.7%), followed by India (57.6%), Sri Lanka (49.7%), and Bangladesh (37.4%), and that for VPA was highest in India (29.9%) followed by Nepal (24.4%), Sri Lanka (24.1%), and Bangladesh (17.9%). Table 2 shows that the prevalence of good SRH was highest in Sri Lanka (73.7%) and lowest in Nepal (37.7%), while in Bangladesh over two-fifth (44.3%) and in India (58.7%) a little less than three-fifth of the population reported being in good health. Results of cross-tabulation also showed that people who reported good SRH were more likely to be in the younger age groups, female, currently married, having no formal education (except for Sri Lanka), self-employed, nonsmoker, and never drinking alcohol. Those who reported engaging in any type of physical activities were also more likely to report being in good health. Table 3. Results indicate that Bangladeshis who engaged in 1-2, 3-4, 5-6, and 7 days a week were respectively 31% [AOR=1. South Asia. Within subgroups, variations were observed in PA. Findings showed that participants from Bangladesh had the lowest proportion of engaging in any type of PA. Similar results on low prevalence of PA were reported by previous studies from Bangladesh 33 and India 34 . Findings suggest that the percentage of good SRH decreased with age in all countries. Female participants were more likely to report good SRH compared to/with males in all countries except for in India. A noticeable variation was observed in the prevalence of SRH among the four countries. Another important disparity was that having higher frequency of participation in PA did not always relate to higher rate of good SRH. For instance, compared to Sri Lanka, participants from Nepal had lower rates of reporting good SRH despite their higher frequency of involvement in both VPA and MPA. A possible connection might be higher living standards of Sri Lanka compared to the other countries measured in terms of Human Development Index (HDI). The correlation between higher educational status and better health outcomes are well documented across countries 18,19 . Our results further indicate that the rate of both moderate and vigorous type physical inactivity was highest in Bangladesh, followed by Sri Lanka, and India. Surprisingly, Nepal had lowest SRH despite its highest prevalence of MPA and second highest prevalence of VPA. As expected, engaging in regular PA was associated with higher odds of good SRH for most of the countries. In Bangladesh and Nepal, those who participated in MPA on a daily basis, and in Sri Lanka, those took MPA 5-6 days a week, had the highest odds of reporting good SRH. For VPA, highest odds of reporting good SRH were reported among those who exercised on daily basis, compared with those who exercised at a lesser frequency. Among all the countries, the strongest associations between SRH and PA in both categories was observed in Sri Lanka.

Results of multivariable regression are shown in
A major barrier to reporting association between SRH and PA is the lack of comparable studies reporting prevalence at the national level and the absence of standardised and validated instruments in studied countries 24 . Previous studies based on USA (86.2%) 20 and Canada (89.9%) 21 concluded that participants who rated their health as poor to average were less likely to take PA compared with those who rated their general health as good to excellent 21 . Similar findings were observed in South Korea, where an independent association between lower level of PA and poor SRH was reported 22 . Regarding the prevalence of PA, a study encompassing 76 countries reported that the prevalence of physical inactivity among individuals aged 15 years or older ranged from 3 to 62% 23 , which varied substantially from the worldwide prevalence of physical inactivity in adults of 31% 24 .
This study has some important limitations. Number of days of VPA and MPA (at least 10 minutes at a time) was used as a proxy for level of PA instead of exact duration. However, similar methodology was used in some other studies 21,26 . Another limitation is the absence of several necessary covariates which are commonly correlated with the level of PA, such as presence of disease conditions (diseased people are less likely to engage in PA), place of residency (urban and rural residences have differing patterns of engaging in PA), and other community level variables (e.g. neighbourhood cleanliness, safety, availability of public spaces for exercise). As the data were secondary, we had no control over the choice of selecting the covariates and the ways they were measured. For instance, we could not distinguish between work and leisure physical activity, which could have affected the associations. With the data being self-reported, there remains a possibility of underand over-reporting, as well as the recall bias. Moreover, there are also differences in the way in which men and women describe their health. Women are more likely to report poorer functioning and worse overall health than men 27 . Last but not least, the results cannot be generalised to all age groups, since no data were available on participants below 18 years of age.